One of the most compelling challenges facing Western democracies is how to maintain and strengthen the bonds of community in ethnically diverse societies. How can we reconcile growing levels of multicultural diversity and the sense of a common identity which sustains the norms of mutual support, the capacity to pursue collective projects and social solidarity?Footnote 1
In much of the Western world, and particularly in Europe, there is a widespread perception that we are failing this test. We are witnessing a backlash against immigration in many countries. There is also a strong sense that multiculturalism policies have “failed,” a reaction that is strongest perhaps in the Netherlands but is felt in many other countries as well. This disenchantment with multiculturalism is driven by fears about economic costs, perceived threats to liberal values, challenges to historic cultures, anxieties about Islam and fears about security. But the reaction is also fueled by a fear that immigration and ethnic diversity are eroding social solidarity. Greater ethnic and racial diversity, some commentators worry, is fragmenting the historic coalitions that built the welfare state. The majority population, they continue, is withdrawing its support from social programs that redistribute resources to newcomers who are seen as “strangers” and not part of “us.” And voters who might otherwise support left-leaning parties are switching their allegiance to conservative parties that oppose immigration, indirectly contributing to a retrenchment in the welfare state they may not have sought in the voting booth.
A tension between diversity and solidarity has been called the progressive's dilemma (Goodhart, Reference Goodhart2004; Pearce, Reference Pearce2004). Many progressives fear they face a trade-off between support for multiculturalism on one hand and support for redistribution on the other. Historically, challenges to immigration and multicultural conceptions of citizenship have tended to come from conservatives committed to preserving historic traditions. Now, doubts are also emerging from the left and centre-left, which increasingly fear that multiculturalism makes it more difficult to advance the agenda of economic redistribution (Cuperus et al., Reference Cuperus, Duffek and Kandel2003).
If these fears are justified, a multicultural welfare state would seem to be in jeopardy. The decline of such in a system might take several forms. Social programs, especially those important to newcomers, might be weakened. Or social programs might be preserved, but immigrants excluded from them through longer residency requirements, a phenomenon often referred to as “welfare chauvinism.” Or national borders might be closed to further immigration, rejecting diversity in the name of protecting the welfare state. Such fears are not merely theoretical conjectures. During 2010, for example, elections in the Netherlands and Sweden—two countries long thought as defining the essence of tolerance—saw populist anti-immigrant parties break through to seize the balance of power. Both of these parties claimed to be protecting the welfare state. During the Dutch election, the leader of the radical right Party for Freedom published a dramatic estimate of social spending on immigrants and demanded that the country “stop the immigration of people from Islamic countries, curtail other immigrant and asylum streams, and exclude new immigrants from welfare benefits for 10 years” (Deutsche Presse Agentur, 2010). During the Swedish election, the Swedish Democrats ran a television advertisement showing an elderly Swedish woman trying to reach an emergency brake labelled “immigration” before a group of burqa-clad women pushing strollers could get to another sign saying “pensions” (Ritter, Reference Ritter2010).Footnote 2
Are such tensions inevitable in diverse democracies? Do progressives everywhere face such a dilemma? If not, what are the factors that mediate between ethnic diversity and solidarity, tipping the balance between inclusive or corrosive relationships in particular contexts? Cross-national research on these issues points in different directions (compare, for example, Alesina and Glaeser, Reference Alesina and Glaeser2004 with Soroka et al., Reference Soroka, Banting, Johnston, Bardham, Bowles and Wallerstein2006). But while large-N comparative studies are invaluable in testing the overly sweeping assertions which permeate this debate, we also need theoretically compelling case studies that dig down to identify the factors that mediate between diversity and solidarity.
This article searches for evidence of the progressive's dilemma in Canada, which clearly is a potentially rich case. Canada is one of the most multicultural countries in the world (Fearon, Reference Fearon2003). If the progressive's dilemma is inherent in the politics of diverse democracies, we should presumably find echoes of the tensions here.
The paper advances a two-step argument centred at the intersection of attitudinal and institutional analysis. Public attitudes in Canada reveal remarkably little tension between ethnic diversity and support for social programs, and the trajectory of attitudinal change does not raise red flags. From this perspective, Canadian experience casts doubt on the idea that a progressives' dilemma is inherent in the politics of diverse democracies. However, understanding why these attitudes prevail in Canada requires an analysis of the wider institutional context within which public attitudes evolve. Institutional analysis tells a different, more contingent story, in which the Canadian policy regime has helped prevent the emergence of a full-blown progressive's dilemma. From this perspective, the Canadian case does not disprove the progressive's dilemma. Rather, Canada has managed to sidestep the dilemma, by diverting adjustment pressures from the welfare state, absorbing some of them in other parts of our policy regime and displacing some of them onto immigrants themselves. Moreover, a focus on institutional mediators does raise red flags for the future. Important protective elements in our policy structures are wearing thin, potentially increasing immigrants' political exposure in the years to come.
This argument is developed in four sections. The first section roots the discussion in different theoretical approaches to the welfare state. The second section briefly summarizes existing evidence on the relationship between diversity and public support for redistribution in Canada. The third section presents a more detailed analysis of the Canadian policy regime, and its implications for a progressive's dilemma in Canada. The final section pulls the threads together and reflects on the durability of the Canadian equilibrium in the future.
Theories of the Welfare State: Attitudes and Institutions
Two distinctive theoretical approaches shed light on the prospects for the welfare state in diverse societies. Both start from the premise that welfare states have been relatively resilient in the face of direct political assaults over the last three decades and seek to explain why. The first approach focuses on public attitudes, explaining the durability of welfare states by emphasizing the embedded nature of public preferences and the slowness with which they evolve. The second approach explains the resilience of welfare states by focusing on institutional dynamics, exploring how complex institutions evolve over time, the ways in which they shape public perceptions and the ways in which they condition the scope for political elites to mobilize citizens' predispositions. Each of these approaches on its own has limits but in combination they take us a long way.
The role of public preferences in shaping public policy has been central to democratic theory from its beginnings. Whatever the conception of democracy, the assumption is that public attitudes matter. Students of democratic representation explore the responsiveness of governments by analyzing the extent to which changes in public preferences cause, or at least precede, changes in public policy (Soroka and Wlezien, Reference Soroka and Wlezien2010; Page and Shapiro, Reference Page and Shapiro1983). From this perspective, the persistence of welfare state depends critically on the stability of supportive mass attitudes (Kumlin, Reference Kumlin, Dalton and Klingemann2007). In their book entitled Why Welfare States Persist, Brooks and Manza (Reference Brooks and Manza2007) argue that welfare states endure because the policy preferences of the public become embedded in the country's social structure, major institutions and the collective memory of citizens. Such attitudes are characterized by remarkable inertia, evolving slowly over generations and constraining radical change in the short term.
From this perspective, the viability of a multicultural welfare state depends on whether increasing ethnic and racial diversity will erode embedded support for redistribution over time. Why would ethnic and racial diversity have such an effect? What are the underlying causal mechanisms? Three theoretical answers have been suggested, each rooted in different disciplines and traditions (see Freeman, Reference Freeman2009, for a useful summary). Theories rooted in evolutionary biology suggest that humans have a natural instinct to be less altruistic towards others with whom they share fewer genes (for example, Salter, Reference Salter2004). Theories based in socio-psychology focus on the role of group identities, arguing that people are comfortable trusting and supporting members of their in-group but tend to be suspicious of members of out-groups. A variant on this interpretation comes from Robert Putnam (Reference Putnam2007), who concludes that in the United States ethnic and racial diversity erodes trust, not just in members of ethnic out-groups but also in members of one's own ethnic group, leading people living in ethnically diverse parts of the country to “hunker down” in social isolation and lose faith in government solutions to social problems. Finally, theories rooted in rational choice emphasize the reciprocal nature of altruism. According to this approach, people are willing to assist and co-operate with people who have assisted and co-operated with them in the past. Each person must be confident others will reciprocate when the time comes, and will not abuse public support (see Fong et al., Reference Fong, Bowles, Gintis, Kolm and Ythier2004). Researchers working in this tradition in the US and Europe find that voters who believe that welfare recipients are simply malingerers tend to be much less supportive of welfare programs and that such stereotypes are especially lethal if they are racialized, with racial minorities or newcomers being seen as heavily dependent on welfare (for example, Boeri, Reference Boeri2009; Boeri et al., Reference Boeri, Hanson and McCormick2002; Gilens, Reference Gilens1999).
Analysts working in all of these traditions accept that features of the wider social context can either trigger or mute these causal mechanisms. The search is on, therefore, for the secret to forestalling such tensions. Most observers place their faith in some form of cultural glue, some form of overarching identity or sense of community which transcends ethnic and racial differences. This faith in community as underpinning of the welfare state has deep roots. The classic statement came from T.H. Marshall, writing about British experience at the end of the 1940s. “Citizenship,” he wrote in an oft-quoted passage, “requires a bond of a different kind, a direct sense of community membership based on loyalty to a civilization that is a common possession” (Marshall, Reference Marshall, Marshall and Bottomore1950: 8). Contemporary analysts echo his view. To offset the corrosive impact of ethnic diversity in the United States, Robert Putnam calls for “a new more capacious sense of ‘we’” (Putnam, Reference Putnam2007: 164). Others look to nationalism as the source of cultural glue. Traditionally, nationalism has been seen as an exclusionary force in history, the source of oppression of marginal groups. However, liberal nationalists argue that a civic conception of national identity can bind together people otherwise divided by economic and ethnic differences (Barry, Reference Barry1991; Miller, Reference Miller1995, Reference Miller, Banting and Kymlicka2006; Rorty, Reference Rorty1998; Tamir, Reference Tamir1993).
But where does such cultural glue come from? Public attitudes are not naturally existing elements in nature. They do not emerge as if by magic. What are the factors that shape citizen preferences? It is at this stage that a purely attitudinal approach tends to run out of steam. To understand the factors that nurture the cultural glue which Marshall and his successors seek, we need to extend the analysis. There are undoubtedly a number of possible approaches at this point, but institutional analysis seems to offer the greatest potential.
The new institutionalism provides a different interpretation of the persistence of mature welfare states. Scholars such as Paul Pierson, Kathleen Thelen, Peter Hall and Jacob Hacker do not look to cultural glue. Rather, they focus on complex institutional dynamics, exploring the ways in which such institutions change through processes of path dependency, punctuated equilibrium, conversion and drift. This approach emphasizes that social programs—once established—exercise feedback effects on the attitudes of the public towards the welfare state. Thus, in explaining the resilience of welfare states in the face of the Reagan and Thatcher governments, Paul Pierson (Reference Pierson1994) pointed to the ways in welfare states generate their own supportive constituencies among the providers and beneficiaries of social programs. Other scholars have examined how the design of specific programs affects individuals' calculus of their self-interest and their sense of the fairness of the system as a whole. They conclude, for example, that selective benefits seem more vulnerable to challenges to the legitimacy of groups of recipients, while universal benefits dampen discussion of which recipients are deserving and which are not (Larsen, Reference Larsen2006; Swank and Betz, Reference Swank and Betz2003; van Oorschot, 2000). Others argue that support for redistribution depends more on trust in government than interpersonal trust, and that such trust is sustained by fairness and effectiveness in the actual administration of government programs (Kumlin and Rothstein, Reference Kumlin and Rothsteinforthcoming; Rothstein, Reference Rothstein1998).
Policy feedback is conceived as including both the ways in which programs distribute material and political resources across different groups and the larger interpretive implications about politics and the welfare state implicit in their structure and operation (Kumlin, Reference Kumlin, Dalton and Klingemann2007; Pierson, Reference Pierson1993). More specifically, policy regimes influence public preferences in three ways. First, policies may generate negative feedback from the public, triggering direct changes in policy, as in thermostatic models of policy change (Soroka and Wlezien, Reference Soroka and Wlezien2010). Second, the structure of programs can alter the nature of information flows to the public; for example, programs which are targeted to particular groups of beneficiaries make their dependency more visible, while universal programs make them less visible. Third, policies—especially policies celebrating cultural motifs and symbols—may influence the sense of national identity held by the public, shaping the collective memories emphasized by Brooks and Manza (Reference Brooks and Manza2007).
This approach offers important insights into the factors shaping Canadian attitudes to diversity and redistribution. In the words of Miriam Smith, “Instead of treating diversity as an independent variable that affects the welfare state, we can invert the question to ask how the welfare state shapes diversity” (Reference Smith2009: 837; also Harell and Stolle Reference Harell and Stolle2010). The scope of the analysis needs to be broadened beyond the welfare state alone to embrace the entire incorporation regime, including immigration policy and integration policies. In combination, these policies set a frame that helps shape public attitudes and expands or narrows the opportunities for political elites to mobilize public attitudes and inject them into political debates.
Public Attitudes in Canada
Over the last few years, our understanding of the relationships between ethnic diversity and public attitudes towards redistribution in Canada has increased considerably. The findings flowing from recent research tend to be reassuring: there is strikingly little evidence of a progressive's dilemma in Canada.Footnote 3
Let's start with attitudes towards immigration and immigrants. Over time, support for existing levels of immigration has grown stronger in Canada (Figure 1). Moreover, in comparison with other leading OECD countries, support for immigration and immigrants is strikingly positive. Table 1 indicates that Canadians are much more likely to believe that immigrants are good for the economy, much less likely to believe that immigrants cause crime and much less committed to reducing immigration. The only dimension in this table on which attitudes parallel those elsewhere is that a substantial majority believe immigrants should adapt to the wider society rather than preserve their own cultural traditions (see also Adams Reference Adams2007; Berry and Kalin Reference Berry and Kalin1995; Harell Reference Harell2009; Wilkes et al., Reference Wilkes, Guppy and Farris2008).
Figure 1 Public Opinion on Immigration, 1988–2006
Source: Canadian Election Study, various years.
Table 1 Attitudes to Immigration and Immigrants
Source: 2003 International Social Survey Program
Analysis of the relationship between ethnic diversity and support for social programs also finds comparatively little tension. Attitudes towards redistribution do not differ strongly across ethnic groups in Canada. When controls for economic and political factors are added to the analysis, visible minority respondents appear somewhat less supportive of universality in health care but somewhat more supportive of redistribution generally, belying occasional suggestions that immigration is bringing a broadly more conservative cast of mind to the politics of social policy. But the differences are not large. Moreover, there is no evidence of majorities turning away from redistribution because some of the beneficiaries are “strangers” (Soroka et al., Reference Soroka, Johnston, Banting, Kay and Johnston2007; Soroka et al., Reference Soroka, Johnston, Kymlicka and Banting2010).
The story remains much the same as we dig deeper, examining factors which analysts elsewhere invoke to explain why diversity erodes solidarity, such as interpersonal trust and perceptions of immigrant dependence on welfare. As in the United States, interpersonal trust declines in ethnically diverse neighbourhoods. But support for social programs does not. Support for redistribution is resistant to the effects of local diversity in Canada (Soroka et al., Reference Soroka, Johnston, Banting, Kay and Johnston2007). An even more striking story emerges around immigrant reliance on welfare. As in other countries, Canadians can be a judgmental people if they suspect free-riding. But perceptions that immigrants are heavily dependent on welfare evoke a more complicated response. While respondents who believe that immigrants are heavily dependent on welfare tend to more critical of social assistance itself, they are more positive about redistribution generally and about other parts of the welfare state. The perception that immigrants are in economic trouble appears to move respondents towards rather than away from support for redistribution generally (Banting et al., Reference Banting, Soroka and Koning2010).
The story remains reassuring when attention shifts to factors that may help to sustain redistribution in diverse societies. Consistent with Rothstein's argument, trust in government is positively associated with support for health care and pensions and, unlike interpersonal trust, trust in government is not lower in ethnically diverse neighbourhoods (Soroka et al., Reference Soroka, Johnston, Banting, Kay and Johnston2007). Even more striking is the role of national identity. In many countries, the arguments advanced by liberal nationalists have seemed like a stretch. The evidence often suggests that individuals with the strongest sense of national identity are most opposed to immigration, and several studies fail to find a positive relationship between the strength of national identity and support for the welfare state (Citrin et al., Reference Citrin, Wong, Duff, Ashmore, Jussin and Wilder2001; Martinez-Herrera, Reference Martinez-Herrera2004; Shayo, Reference Shayo2009). In Canada, however, research finds partial support for liberal nationalist arguments. Those with the strongest sense of Canadian identity embrace immigration and immigrants more warmly than their less nationalist neighbours. Since anti-welfare sentiments are associated with opposition to redistribution, a strong sense of national identity clearly offsets this, helping to insulate the welfare state from the toxic effects of anti-immigrant sentiment (Johnston et al., Reference Johnston, Banting, Kymlicka and Soroka2010).Footnote 4
All of this suggests that Canada has avoided the progressive's dilemma. Moreover, while attitudes towards immigration vary with the economic cycle, weakening during recessions, there seems to be no long-term trend towards greater resistance to immigration. Viewed from this perspective, the country seems to have found a sustainable equilibrium among support for immigration, multiculturalism and the welfare state. One might therefore be tempted to declare the progressive's dilemma to be a myth, an invention of anxious Europeans going through an evolution which societies like Canada managed long ago.
But why? Resorting to an argument about Canadian exceptionalism seems unsatisfying. What is it about the Canadian context that facilitates this equilibrium? The next section extends the search to the institutional level, focusing on the Canadian incorporation regime.
The Canadian Incorporation Regime
The Canadian incorporation regime consists of policies designed to facilitate the economic, social and political incorporation of newcomers and minorities (Breton, Reference Breton and Reitz2005; Myles and St-Arnaud, Reference Myles, St-Arnaud, Banting and Kymlicka2006). The key components of the regime—immigration policy, integration policy and social policy—interlock but do not mesh perfectly. As Freeman comments, “No state possesses a truly coherent incorporation regime…. Rather one finds sub-system frameworks that are weakly, if at all, co-ordinated” (Freeman, Reference Freeman2004: 946). In the Canadian case, there is constant friction among the components; they change at different speeds in response to different pressures, often impacting on other components of the model in unexpected ways. But, in combination, they matter.
Immigration Policy
Canadian immigration policy has been carefully designed to forestall heavy immigrant reliance on social benefits, which has triggered backlash elsewhere. Since the adoption of the points system, Canada has given priority to economic migrants who are capable of moving quickly into employment and self-sufficiency. The contrast with Europe is sharp. The early postwar waves of immigrants to Europe countries represented a post-colonial legacy or guest workers recruited as temporary occupants of low-skilled jobs. The lingering consequences for the employment prospects of immigrants compared to native-born workers are captured in Table 2.Footnote 5 High immigrant unemployment in European countries has inevitably led to greater reliance on social benefits, fueling populist fires.
Table 2 Unemployment among Native-Born and Foreign-Born, 2004
Source: Calculated from data in OECD, 2006b, Table 1.A1.2. Data for Canada refer to 2003.
Other elements of immigration policy have also been managed to minimize reliance on welfare. Immigrants wishing to sponsor a family member to come to the country must sign an “undertaking” with the state, promising to provide financial support for the family members they sponsor for a period ranging from three to ten years (Côté et al., Reference Côté, Kérist and Côté2001). During that period, family members are ineligible for social assistance. In the case of refugee claimants, Canada takes a step almost unheard of elsewhere, extending temporary work permits to refugee claimants if they would otherwise have to turn to public assistance. In 2003–2005, for example, 76 per cent of adult refugee claimants had a temporary work permit (Yu et al., Reference Yu, Ouellet and Warmington2007: 25). In addition, the ability of temporary foreign workers to access benefits is constrained in multiple ways (Koning and Banting, Reference Koning and Banting2010). The overall pattern is clear: the immigration regime has been designed as much as possible to limit immigrant utilization of welfare.
Moreover, immigration policy has been adjusted from time to time to block any trend towards reliance on welfare, as the 1990s illustrated. That decade saw the emergence of the Reform party, which temporarily disrupted partisan consensus on immigration, as well as a deep recession and large government deficits. Opinion polls pointed to a hardening of public attitudes both towards social programs and immigration (Abu-Laban, Reference Abu-Laban1998; Abu-Laban and Gabriel, Reference Abu-Laban and Gabriel2002). In this environment, the newly elected federal Liberal government launched a national consultation on immigration policy. A recurring theme through the consultation was anxiety about the burden of immigrants on social programs, especially family-class members admitted under a sponsorship agreement. An internal federal study concluded that 14 per cent of immigrant families in Toronto who were covered by sponsorship agreements were actually on social assistance in 1993, and estimated that the cost on a national level basis was $700 million each year.Footnote 6 A report commissioned for the consultation concluded that “public confidence in the administration of the family-class program has been shaken” and that “in practice, the sponsorship agreement system is in tatters” (Hathaway, Reference Hathaway, papers and papers1994: 14–15; also Citizenship and Immigration, 1994a: 11).
In many European countries as well as Australia and the United States, such politics led to the adoption of longer residency periods for welfare programs. But the Canadian federal government does not control the terms of access to social assistance and health care, which are provincial responsibilities.Footnote 7 It therefore pulled the lever it does control, admission criteria. The government announced that it would shift the balance between economic immigrants and family-class immigrants and was explicit about the rationale: “The relationship between the main components of immigration will be shifted to focus more on those less likely to require public assistance. Over time, economic migrants will compromise a larger percentage of newcomers as compared to family immigration” (Citizenship and Immigration 1994a: 62). The proportion of family-class admissions fell like a stone, dropping from 51 to 30 per cent of the total by 1996. As Figure 2 makes clear, the dominance of the economic class has remained unchallenged since.
Figure 2 Composition of Immigrant Inflow by Category of Immigrant, 1984–2008
Source: Koning and Banting (Reference Koning and Banting2010)
Canadian immigration policy can be justifiably criticized from the perspective of global social justice. We have absorbed considerable talent from poorer counties. But from the Canadian perspective, this strategy has dampened one of dynamics that has led to backlash elsewhere. Throughout the 1990s and early 2000s, study after study concluded that immigrants to Canada relied less on income transfers than the native-born population (Baker and Benjamin, Reference Baker and Benjamin1995a, Reference Baker, Benjamin and DeVoretz1995b; Sweetman, Reference Sweetman, Schwartz and Aydemir2001). This made it difficult for critics to sustain a dominant political narrative of gloom about the economic and social costs of immigration. It undoubtedly has shaped Canadian perceptions of immigrants as hard working and contributed to the paradoxical finding discussed earlier that those who see immigrants as most heavily dependent on welfare are more supportive of redistribution generally.
More recently, the political buffering effects of immigration policy are starting to wear thin. The economic incorporation of newcomers has slowed dramatically in recent years. Newcomers have higher levels of education and training than previous cohorts of immigrants, yet they are entering the workforce at much lower income levels and taking much longer to close in on the income of the native-born population.Footnote 8 As a result, immigrants' need for income support is increasing. The shift is less marked than it might be, as the economic problems of immigrants often take the form of underemployment rather than outright unemployment. The pain and frustration of underemployment is real, but it is born by the immigrant families themselves. Despite this, the need for income support has grown.
Figure 3 tracks utilization of two relevant programs, Employment Insurance and social assistance.Footnote 9 The slowing economic incorporation of recent cohorts is increasing the need for support. In the last few years, support has nudged marginally above the level of non-immigrants. The difference at the national level is small; in regions where most immigrants reside, the gap may be larger. Clearly, the insulation generated by immigration policy has grown thinner, and Canada seems to be moving into territory that has proven politically treacherous elsewhere.
Figure 3 Combined Social Assistance and UI/EI Transfers, Per Capita by Immigration Status, 1982–2007
Source: Calculated from Statistics Canada, Survey of Consumer Finances (1982–1998) and Survey of Labour and Income Dynamics (1999–2007). Data supplied by Edward Koning.
Integration Policy
As we shall see, Canadian integration policy has also supported a multicultural welfare state.Footnote 10 It is worth pausing, however, to note that this conclusion dissents from assertions elsewhere that multiculturalism policies generate a politics of interethnic relations that undercuts support for redistribution (Barry, Reference Barry2001; Rorty Reference Rorty1998, Reference Rorty2000; Wolfe and Klausen, Reference Wolfe and Klausen1997, Reference Wolfe and Klausen2000). There is no evidence that this is a consistent phenomenon; cross-national analysis demonstrates that countries with strong multicultural policies do not, on average, have greater difficulty sustaining their welfare states (Banting et al., Reference Banting, Johnston, Kymlicka, Soroka, Banting and Kymlicka2006). Nevertheless, critics in individual countries insist that multiculturalism exacerbates the progressive's dilemma. For example, Koopmans (Reference Koopmans2010) argues that in the Netherlands and other European countries, multiculturalism increases social segregation, slows language acquisition, weakens the educational attainments of the second generation and limits economic integration. When combined with a generous welfare state, the result is high levels of welfare dependency.
This critique cannot be applied to Canada. It reflects neither the intent nor the consequences of the Canadian approach. The intention underlying multiculturalism policies in this country has always been integrationist (Kymlicka, Reference Kymlicka1998; Ley, Reference Ley, Vertovec and Wessendorf2010). The core goals of federal multiculturalism were laid out in 1971: recognition and accommodation of cultural diversity; removing barriers to participation; promoting interchange between groups; and promoting acquisition of the official languages. Changes in the federal program over the years have reinforced the integrative intent. Starting in the mid-1970s, the goals of participation and interethnic exchange increasingly received the lion's share of funding under the federal program. Interestingly, when people talk about the “retreat from multiculturalism” in the Netherlands, they often cite the decision to shift funding away from monoethnic organizations to multiethnic ones that promote cultural interchange and co-operation across ethnic lines. This shift had taken place already in Canada by the late 1970s, where it was understood as an evolution guided by the original goals of multiculturalism, not as a “retreat” from them.Footnote 11
Nor have multiculturalism policies been segregrationist in their impact. Critics often point to residential segregation, which is increasing (Hou, Reference Hou2004). But the number of significant ethnic “enclaves” is still limited (Myles and Hou, Reference Myles and Hou2004; Walks and Bourne, Reference Walks and Bourne2006). More importantly, residential concentration is not driven by multiculturalism policies. Such concentrations are common in countries which have rejected the multicultural philosophy, and in Canada, current concentration levels remain lower than the concentration of earlier European immigrant groups who arrived long before multiculturalism policies were adopted (Hou, Reference Hou2004). Nor are our multiculturalism policies weakening language acquisition or the educational attainments of immigrant children. Indeed, the opposite is the case. Canada has a strong comparative advantage in educating immigrant students, in part, according to the OECD, because of specific policies to address issues of cultural and linguistic diversity in the school population (OECD, 2006a). These diversity policies help to explain why the children of immigrants do better in Canada even when one controls for the skills, education and income of their parents. Similar patterns emerge in political integration. Irene Bloemraad's careful comparison of the political integration of immigrants in Canada and the United States demonstrates that multicultural programs encourage and enable newcomers to participate more quickly and more effectively in mainstream Canadian institutions (Bloemraad, Reference Bloemraad2006).
The multiculturalism strategy has also had an impact in nurturing a more inclusive sense of identity which liberal nationalists see as sustaining a redistributive state. The adoption of the policies of bilingualism and multiculturalism in the 1960s and 1970s represented a state-led redefinition of national identity, an effort to de-centre the historic conception of the country as a British society and to build an identity more accommodating of Canada's cultural complexity (Champion, Reference Champion2010; Igartua, Reference Igartua2006; Uberoi, Reference Uberoi2008). As Harell (Reference Harell2009) observes, multicultural norms have helped to “normalize” diversity, especially for younger generations, slowly reshaping the embedded collective memories, which Brooks and Manza (Reference Brooks and Manza2007) see as part of the attitudinal foundations of the welfare state. The result is a conception of national identity which strengthens support for the welfare state in a diverse society; as we have seen, unlike in most countries, those with strongest sense of Canadian identity tend to be more sympathetic to immigrants, and that sentiment helps mitigate the toxic effects that anti-immigrant sentiment would otherwise have for redistribution.
Although the integration component of the Canadian incorporation regime has changed least in recent years, multiculturalism is hardly without challengers (Ryan, Reference Ryan2010). As we saw in Table 1, Canadians have a strong preference for immigrants to adapt to Canadian society, and some Canadians worry that immigrants are not adopting Canadian norms. Many Québécois commentators see federal multiculturalism as a denial of the centrality of dualism in Canada (Gagnon and Iacovino, Reference Gagnon and Iacovino2007). And aspects of the Reform party's opposition to multiculturalism undoubtedly linger in the current federal government. Nevertheless, the concept of multiculturalism is embedded symbolically in the Canadian constitution, and powerful pressures for change have not yet been mobilized. The danger is that this component of the incorporation regime may be forced to carry greater weight as a result of shifts in the other two.
Social Policy
In part, the welfare state is the author of its own fate. The structure of social policy influences the politics of social policy, including attitudes about redistribution and the prospects for backlash. In Canada, these processes have traditionally worked to forestall a progressive's dilemma, but this protection is getting thinner.
Compared to many countries, Canada has not sought to fence off the welfare state from newcomers with a formidable set of residency requirements. In many countries, including other settler societies, immigrants face waiting periods of two to five to ten years for access to important benefits, and the length of the wait has been growing (see Carney and Bouchard, Reference Carney and Bouchard2010). In Canada, explicit restrictions on the access of immigrants to social programs are limited. Immigrants who have become permanent residents have immediate, full and equal access to almost all social programs. There are important exceptions for some programs and some groups, most notably, recently arrived elderly immigrants, family-class immigrants under sponsorship agreements and temporary foreign workers (Koning and Banting, Reference Koning and Banting2010). But compared to welfare states elsewhere, Canadian social policy does not keep newcomers at bay for long. In part, of course, this is because the employment-focused nature of immigration policy has provided an alternative mechanism for minimizing welfare dependency. It is revealing that the most stringent residency requirement is for Old Age Security, where an employment strategy is clearly less relevant (Baker et al., Reference Baker, Benjamin and Fan2009).
Does the modest nature of Canadian social programs explain the modest level of welfare chauvinism? Canada has long been classified as a “liberal” welfare state (Esping-Andersen, Reference Esping-Andersen1990). Compared with European social models, the Canadian income floor is defined by lower benefit levels and shorter benefit duration. Canada spends less on income transfers such as unemployment benefits and family benefits than most OECD countries, and partly as a result, taxes and transfers do not reduce inequality as much as in the more ambitious welfare states (OECD, 2008). The liberal nature of the Canadian welfare state was established long before the advent of modern forms of diversity. Nevertheless, Europeans might be forgiven for wondering whether Canadians have avoided the progressive's dilemma by not being very progressive. In support of such suspicions, they might point to an interesting exchange between the federal minister of immigration and the Ontario minister responsible for social services. The federal minister complained to provincial welfare ministers that the “generosity of our social welfare schemes” encourages foreigners to submit bogus refugee claims in order to live off welfare for a few years. The Ontario minister's reply was interesting: “I wouldn't say that our welfare system encourages fraudulent refugee claims… I don't think [welfare] is generous anywhere in the country” (Mahoney Reference Mahoney2010).
Nevertheless, Canada's status as a liberal welfare state is not the critical factor. The progressive's dilemma has been a powerful focus of debate in other liberal welfare states: the term was coined as a description of the UK experience; the United States is a frequently quoted example; and Australia has established more extensive residency requirements.
The structure of social programs seems to matter more. As we have seen, analysts have repeatedly concluded that universal programs dampen welfare chauvinism, whereas means-tested programs are more vulnerable. A careful comparison of Australian and Canadian programs indicates that Australia relies more heavily on residency requirements largely because its core social programs are means-tested (Carney and Bouchard, Reference Carney and Bouchard2010). Certainly, in Canada, universal programs attract less political controversy about immigrant utilization. Access to health care for permanent residents seems uncontested. Social insurance programs, such as the Canada/Quebec Pension Plans and Employment Insurance, which depend on employment and contributions, spark no excitement. Social assistance, in contrast, attracts political heat. Refugee claimants on welfare can generate sparks, as we have seen. But the largest point of friction concerns family immigrants under sponsorship agreements. Individual sponsorships can break down because marriages or family relationships collapse or because sponsors themselves get into economic difficulty.Footnote 12 Not surprisingly, the numbers grow during recessions, prompting provincial campaigns to crack down on “deadbeat” sponsors. In the 1990s, Quebec began to put pressure on sponsors, including efforts to garnishee their wages (National Council of Welfare, 1997); and Ontario began to deduct $100 per month from the social assistance payments to immigrants whose sponsors had defaulted.Footnote 13 This strategy was challenged by a group of legal clinics as discriminatory under the Charter, and in 2004 Ontario settled the case, stopped reducing welfare payments and joined Quebec in putting pressure on sponsors. Eight sponsors who were forced to repay social assistance debts appealed to the courts to have their debt waived (Mavi v. Canada, 2009). The provincial court ruled that Ontario does not have the right to unilaterally enforce all sponsors to repay the debt, and that instead they are under obligation to decide on a case-by-case basis and are held to standards of procedural fairness. The case was appealed to the Supreme Court of Canada.
Given the greater political exposure of immigrants in social assistance, the restructuring of the welfare state in the 1990s was counterproductive for a diverse society. Although restructuring was not driven by the politics of diversity, there is potential fallout. One of the changes has been a shift in the balance between Employment Insurance to social assistance. Employment Insurance was never a perfect vehicle for immigrants, as it only helps those who have successfully integrated into the labour force, but participation is politically self-legitimating. Since the mid-1990s, however, program cuts and changes in the labour market have reduced the proportion of the unemployed actually receiving unemployment benefits from approximately 80 per cent to 40 per cent. The shift is accentuated for immigrants by the regional nature of the program. Coverage rates are dramatically lower in areas where the percentage of immigrants is higher. In Toronto, where almost one in two residents is an immigrant, only 22.3 per cent of unemployed people was receiving unemployment benefit in 2004 (Battle et al., Reference Battle, Mendelson and Torjman2006). These inequalities reflect the politics of regionalism, not the politics of immigrant diversity. But the reality is that the federal government has largely opted out of income stabilization in half of the country where most immigrants live. In these regions, social assistance is moving from last resort to first resort. Now, when times get tough, immigrants must increasingly turn to social assistance, where they are more politically exposed.
Summary and Prospects
The progressive's dilemma is not a lived reality in Canada today. Public attitudes reveal markedly little evidence of a deep tension between ethnic diversity and support for redistribution. In comparative terms, Canada is an outlier in terms of the status of immigrants in the symbolic ordering of the country, and the legitimacy of immigrants in a society of immigrants buffers the welfare state from the harsher forms of welfare chauvinism. A multicultural identity seems to provide at least some of the cultural glue sought by observers concerned about the progressive's dilemma.
These attitudes, however, are not a spontaneously occurring substance in nature. They are not independent of the gritty world of public policy. Over the last decades, the incorporation regime has nurtured these attitudes and conditioned they ways in which public attitudes are mobilized politically. Immigration policy minimized the dependence of newcomers on social support, providing less room for attacks on immigrants as a burden on the country. Integration policy represented a state-led transition to a multicultural conception of the country, building an identity which has helped protect the welfare state from anti-immigrant sentiments. And universal social programs reduced the exposure of immigrants to the politics of welfare chauvinism. In combination, the incorporation regime helped keep the progressive's dilemma at bay.
Past successes can never be taken for granted. As we have seen, the protection is thinning in critical areas. The slowing economic integration of newcomers has increased their need for support, and their average benefits now exceed those of the native-born. The dangers are heightened by the shift from Employment Insurance to social assistance, especially in parts of the country where most immigrants live. As a result, we seem to be heading towards territory that has proven politically combustible elsewhere. So far, the integration component of the policy regime has been subject to less change, but one wonders whether it can legitimate a multicultural welfare state alone, without supporting buttresses.
The theoretical implications of this analysis are worth highlighting. Both cultural and institutional analyses are required to understand the viability of the welfare state in diverse societies. We need to track the evolution of the attitudinal foundations of the welfare state, and the strength or weakness of cultural glue knitting together a diverse society. But an exclusive reliance on attitudinal explanations has limits. It cannot explain where supportive attitudes come from and why they change over time. Institutional analysis helps here. Policy regimes set a frame which helps shape attitudes, and expands or narrows the opportunities for political elites to mobilize the darker side that exists in any democratic citizenry. Phrased in another way, how much cultural glue a society needs depends not only on the extent and nature of diversity as such, but also on the ways in which public policies structure the distribution of benefits and burdens across a diverse population. In the Canadian case, interlinked policy frameworks forestall the sorts of distributions that can lead to levels of ethnic tensions that would require especially large doses of cultural glue.
The bottom line for Canada can be distilled simply. If the progressive's dilemma does emerge in Canada, if the public turns on immigrant access to social programs or cheers when governments cut programs on which immigrants rely, careful analysis will undoubtedly reveal shifts in public attitudes. But the full story is unlikely to be a Putnam-like story. Backlash is unlikely to emerge because ethnic diversity is finally having its long-predicted corrosive effects on support for redistribution. If we trace back the causal chain, we are likely to find that such a shift had been triggered in the first instance by institutional changes driven by pressures which may well be unrelated to diversity but which increase the opportunities to mobilize the darker side of Canadians' views.
Is there any danger of such a scenario unfolding in the years to come? Immigration policy is currently undergoing dramatic restructuring. As in the past, the Canadian government has reacted to problems of economic integration by changing admissions policy. Immigrant selection is being decentralized and privatized; the points system is less central; the number of temporary foreign workers is increasing dramatically; new pathways to citizenship are being open for some but not all temporary residents (Alboim, Reference Alboim2009). It is much too early to assess the impacts of the new immigration strategy. Perhaps the changes will reinvigorate economic integration, further buffering the welfare state. Perhaps they are a perfect recipe for welfare chauvinism. We cannot know for sure. Nevertheless, it is troubling that these changes are happening in response to immediate pressures, with little public debate about the long-term implications for the incorporation process as a whole.
In Britain, Trevor Phillips, head of Equalities Commission, famously pronounced that Britain was “sleep-walking into segregation.” It is important that Canada not sleep-walk into a progressive's dilemma.
